Generating Multilevel Voronoi Tessellation in DREAM.3D-NX Without Experimental Data

[oaugestad88]

Dear DREAM.3D-NX Support Team,

I am working with DREAM3D-NX 7.0.2 on Linux and want to generate a multilevel Voronoi tessellation microstructure for dual-phase steel entirely from scratch—without importing any experimental data. My goal is to simulate a fully synthetic microstructure with a two-phase system (e.g., ferrite and martensite) within a representative volume element (RVE).

I have several technical questions regarding the best workflow to build this from scratch:

1. Generating a Multilevel Voronoi Tessellation Without Imported Data

   • What is the recommended method to randomly distribute seed points within an RVE for the initial Voronoi tessellation?
   • After generating the first-level Voronoi structure, how can we select a subset of grains and perform a second Voronoi tessellation within them?
   • Which filters and processing steps are required to achieve this multilevel tessellation approach efficiently?
   • What is the best way to assign different material phases to each level of the tessellation to approximate dual-phase steel?

2. Reference Examples and Workflow

   • Are there any predefined microstructures in the Bookmarks that we can use as a reference for building a microstructure from scratch?
   • Is there an example pipeline that demonstrates seed generation, tessellation, and phase assignment without relying on experimental EBSD or imaging data?

3. File Export for DAMASK Compatibility

   • What is the best way to export the generated microstructure in formats compatible with DAMASK (e.g., .vti, .geom)?
   • Does DREAM.3D-NX support direct .vti file export, or is an intermediate conversion required?
   • If .vti files can be written, what parameters should we ensure for compatibility with external FEM solvers like DAMASK?
   • Are there any alternative workflows or scripts available to generate DAMASK-ready microstructure files?

Any insights, documentation, or example pipelines would be greatly appreciated.

Best regards,
Oscar

[imikejackson]

Hello @oaugestad88
DREAM3D-NX version 7 currently does not have any synthetic microstructure generation capabilities. There currently is no time line for adding these capabilities. If you would like to try DREAM.3D version 6.5.171 from https://dream3d.bluequartz.net please feel free to try that version.

[oaugestad88]

Thank you @imikejackson or your response and for pointing me toward DREAM.3D 6.5.171. I have successfully installed it and am now able to generate my own synthetic microstructures. I truly appreciate the support!

As I continue refining my workflow for building a dual-phase (DP) steel microstructure from scratch, I have a few additional questions:

1. Randomizing Euler Angles Without Input Data

   When generating a microstructure without any experimental input data, how can I randomize Euler angles for grains that do not have preset values?
   Is there a recommended filter or approach to ensure a realistic grain orientation distribution in the absence of EBSD-derived data?

2. Voronoi Tessellation & Multilevel Structures

   What is the best method to generate random seeds for Voronoi tessellation within an RVE?
   Once the initial tessellation is created, how can I perform a multilevel Voronoi tessellation, where a subset of grains undergoes further subdivision?
   Are there specific filters or processing steps that facilitate hierarchical grain refinement in a controlled manner?

3. Modifying Example Structures for DP Steel

   I have been working with the Single Cubic Phase Equiaxed example structure and am trying to modify it to represent a dual-phase steel microstructure. What would be the most effective way to introduce two distinct phases within the tessellation?
   Would the "Generate IPF Colors" filter be useful in distinguishing the phases visually, or is there a more appropriate method?

4. Additional Considerations

   Are there any best practices for assigning phase fractions and properties when dealing with multilevel tessellations?
   If I want to simulate different phase distributions, would it be best to control this at the seed generation stage, or should it be applied post-tessellation?

I really appreciate your guidance on these topics and any additional resources or recommendations you may have. Thank you for your time and support!